Optimized bio-inspired water quality level prediction framework

Purpose In recent years, several pollutants have affected the water quality. Consequently, water quality prediction and modeling became the hottest topic. Therefore, the typical networks’ features are limited, so the water quality forecasting results in poor prediction. Design/methodology/approach This paper introduced a novel Lotus-based Lenet Prediction System (LbLPS) as a recognition framework to determine water quality in the Shanmuganadhi River of the Theni district of South India. The water quality data set was initially collected, and the noisy elements were eliminated. The needed features were extracted and analyzed. Lotus fitness follows-up on the feature extraction and classification process. Here, the MATLAB environment functioned in the LbLPS model. Eventually, the proposed LbLPS performance was authenticated. Findings Performance criteria such as F-score, precision, accuracy, recall, computational time and error rate were evaluated. The accomplishment of the introduced model LbLPS obtained higher F-score, precision, accuracy and recall with a meager error value and computation time. Originality/value The accomplishment of the introduced model LbLPS obtained a higher F-score, precision, accuracy and recall with a meager error value and computation time.